Artificial Intelligence Based Estimation of Water Saturation Using Electrical Measurements Data in a Carbonate Reservoir

Abstract

Good estimation of water saturation is necessary for successful estimation of reservoir properties as it minimizes the error in initial oil-in-place calculations. Electrical measurements on core plugs have been used to predict water saturation by analyzing the parameters of Archie's equation. Presently, several methods such as conventional, CAPE (1, m, n), CAPE (a, m, n) and 3D methods have been used to analyze the parameters. However, the accuracy of these methods has become inadequate for optimal estimations. Recent successful applications of Artificial Intelligence (AI) techniques in petroleum engineering have confirmed the capability of AI to handle such non-linear and complex industrial problems. Interestingly, the new paradigm of employing AI for water saturation estimation has not been sufficiently addressed in the petroleum engineering technology application literature.This study focuses on the use of the predictive capabilities of Artificial Neural Networks and the Fuzzy Inference Engine. 378 data samples obtained from the laboratory measurements of electrical properties of 41 core plugs taken from Carbonate reservoir rocks in the Middle East were used for the implementation of the proposed techniques. Using the popular stratified data sampling method, 70% of the data points were used to train the AI models while the remaining 30% were used for validation and testing.The comparative results showed that the AI models performed better with higher accuracy and lower errors than those obtained with the current methods. Based on this result, we conclude that Fuzzy Logic exhibits a robust predictive capability for the estimation of water saturation from electrical measurements by providing a good match with the core values and giving better error distribution than the other AI and current methods.

Machine Learning in Oil and Gas Exploration: A Review

“…The results showed that ANFIS provided slightly better output accuracy than ANN. Meanwhile, ANN and FL were compared by Bageri et al [114] in a carbonate reservoir in the Middle East. The output suggests that the FL model offers higher accuracy than the ANN model.…”

AI-Driven Maintenance Support for Downhole Tools and Electronics Operated in Dynamic Drilling Environments

“…However, in this context the dominant focus is Reservoir modelling & Characterization, e.g. [55]- [57] and Completion & Production related applications, e.g. [58]- [61].…”

Investigating the effect of training–testing data stratification on the performance of soft computing techniques: an experimental study

“…It was used by Bageri, Anifowose, and Abdulraheem (2015), Helmy, Fatai, and Faisal (2010), Imam, Anifowose, and Azad (2015), Khoukhi and Albukhitan (2010), nooruddin, Anifowose, and Abdulraheem (2013), olatunji, Selamat, and raheem (2011), and oloso, Khoukhi, Abdulraheem, and Elshafei (2010). Bageri et al (2015) used the predictive capabilities of Ann and the Fuzzy Inference Engine to predict the water saturation of a Middle Eastern carbonate petroleum reservoir. utilised the capabilities of data mining and CI in the prediction of porosity and permeability based on the hybridisation of fuzzy logic, SVM, and functional networks (Fn).…”